Review of the Properties and Reactions of Some ... - ACS Publications

1

R e v i e w of the

Properties and

Reactions

of S o m e Inorganic Free R a d i c a l s F. O. RICE

Downloaded by 219.148.39.134 on December 19, 2015 | http://pubs.acs.org Publication Date: January 1, 1962 | doi: 10.1021/ba-1962-0036.ch001

Department of Chemistry, Georgetown University, Washington 7, D. C.

The chemical properties and reactions of free radicals are more particularly emphasized, even though the more spectacular recent advances in our understanding of free radicals have come through the use of tools and techniques taken from the domain of the physicist. It seems worth while to explore whether or not the methods of classical chemistry with suitable modifications can be developed further to increase our knowledge of transformations involving free radicals. The stabilization of various radicals and the type of reactions characteristic of divalent radicals are discussed.

This review emphasizes more particularly the chemical properties and reactions of free radicals, even though the more spectacular recent advances in our under standing of free radicals have come through the use of tools and techniques taken from the domain of the physicist. It seems worthwhile to explore whether or not the methods of classical chemistry with suitable modifications can be de veloped further to increase our knowledge of transformations involving free radicals. In general, a univalent free radical is not easily disposed of; if it reacts with surrounding molecules it will itself disappear but will generate another free radi cal. The only way a univalent free radical can disappear is to collide with an other univalent free radical and either combine with it or disproportionate. This result is a necessary consequence of the fact that, with almost negligibly few ex ceptions, chemical molecules contain an even number of electrons. It is clear, too, that the generation of a univalent free radical in a chemical system will start a whole chain of reactions that will end only by collision of two radicals. We can understand this by realizing that the initial act, generating the free radical, since it breaks a bond, must require a high activation energy. The subsequent reactions, while they involve breaking bonds, involve also the con comitant making of bonds and have a relatively small activation energy; it is evi dent therefore that the conditions that initially produce the free radicals will cause the chain reactions to proceed with relatively high speed. The study of these ma terial chains cleared up many problems, notably the inhibition of chemical 3

In FREE RADICALS in Inorganic Chemistry; COLBURN, C.; Advances in Chemistry; American Chemical Society: Washington, DC, 1962.

ADVANCES IN CHEMISTRY SERIES

4

reactions, that plagued chemists during the early part of the nineteenth century. During that period, attempts were made to develop the concept of energy chains, but these attempts failed because they could not explain the specificity of chemical transformations. It is possible that the curious interchange of resonance light between atoms may lead to some developments of an energy chain.


Nitric Oxide Perhaps the best known of the inorganic free radicals is the substance nitric oxide, one of the very few "odd" molecules that are known to exist. Its method of preparation is a typical one. Briefly, an equilibrium or stationary state is established at elevated temperatures and then the temperature is suddenly reduced to a low enough point so that the equilibrium is frozen. It is then possible to study the chemical behavior of the particles at leisure. Among free radicals, nitric oxide is peculiar in that it behaves as a stable molecule at ambient temperatures. Its stability (18) may be interpreted as resulting from resonance between the structures +N—O- and N—0+. In the liquid and solid states, combination occurs and nitric oxide exists entirely or almost entirely as the dimer. It seems probable that trapping NO in a matrix would be a useful way of studying the effectiveness of this method of stabilizing radicals. To prepare nitric oxide a mixture of nitrogen and oxygen may be heated to about 3000° C , at which temperature air contains about 4% of NO at equilib rium; by rapid cooling to room temperature one can obtain 2 to 3% of NO in the air. Ordinarily, in order to stabilize free radicals it is necessary to cool the reaction mixtures to liquid nitrogen or even liquid helium temperatures. The preparation of NO by the oxidation of ammonia over a platinum catalyst suggests that other radicals would be formed in this way. _

Methods of Stabilizing Radicals Free radicals of short life may be produced by heating a substance at low pressures in a flowing system to a high temperature, whereby a portion of the substance is decomposed. The products are rapidly brought in contact with a surface that is cooled to liquid nitrogen or even liquid helium temperatures. This simple method is adequate to prepare active species, as shown by Rice and Freamo's experiments (24) on the preparation of a blue material by the thermal decomposition of hydrazoic acid. Doubtless there were numerous attempts to do this sort of experiment—for example, Staudinger and Kreis (30) report in some detail unsuccessful attempts to prepare S . They used a "hot-cold" quartz tube which contained sulfur at one end and was evacuated and sealed. When the end of the tube containing the sulfur was heated, it distilled to the other end, which was cooled to liquid nitrogen temperatures, but only ordinary yellow sulfur was obtained. However, Dewar and his coworkers (5) actually prepared CS using this technique. Apparently CS in the gas phase has the stability of a normal molecule, but there is a heterogeneous reaction at surfaces, causing polymeriza tion (7,10). We have to ask what happens when a radical or molecule approaches a surface cooled to liquid nitrogen temperatures or lower. Some of the particles undoubtedly bounce off and there is almost certainly some lateral movement of those that stick on the first collision. The problem of immobilizing radicals on their first collision is critical to the solution of the problem of obtaining them in high concentration. This is discussed in the imaginative experiments of Windsor, who attempted to prepare spin-aligned hydrogen atoms (1 ). 2



RICE

5

Inorganic Free Radicals

Very soon after the publication of the work of Rice and Freamo there appeared independently the suggestion that the stabilization of free radicals produced would be facilitated by the addition of some inert material which would dilute the substrate passing through the furnace. Many papers have been pub lished on this technique, which is commonly known as the matrix isolation method (16, 33). However, the substances commonly used to form the matrix are mo lecular solids, so that the forces between the molecules are very weak and con sequently radicals can be preserved only at temperatures near the boiling point of liquid nitrogen or even lower. If one could incorporate radicals into a diamondtype lattice, it might be possible to stabilize radicals sufficiently to keep them at room temperature. Another possibility would be to incorporate radicals into an ionic lattice in which the strong forces existing between the particles could be expected to prevent diffusion. In some experiments along these lines (27), in which hydrazoic acid was mixed with vapors of alkali halides, the mixture being passed through a furnace at 800° to 900° C. and a pressure of a few tenths of a millimeter, difficulties were encountered due to formation of /-centers. If the alkali metal and the halogen are not present in exactly stoichiometric proportions, there are "holes" in the lattice, producing negative or positive charges and color formation, so it is not certain whether colors are due to radicals or /-centers. Because of these difficul ties the work was discontinued. Chilton and Porter (3) have published a research note on the stabilization of free radicals in salt matrices, but their method of preparation would seem to indicate that they simply had microcrystals of a sub strate embedded in the salt matrix. An ideal matrix would consist of a nonpolar molecular solid to minimize interaction between the free radical species and the matrix molecule. To study a free radical by physical methods it is desirable to have the conditions as nearly as possible those existing in a dilute gas. Another possible method of stabilizing free radicals would be to imprison them in clathrate compounds (14, 23). A clathrate is a homogeneous solid consisting of a host and a guest molecule; the guest is imprisoned in closed cavities or cages formed by the host molecules. If it were possible to find a suit able stable host molecule that would form cages in the matrix on freezing from the dilute gas phase, it should be possible to carry with the vapor free radicals which should then be stabilized in their cages. An ingenious method of stabilizing free radicals is that of making a "fluff" of an unsaturated polymer such as polystyrene (11). In this way radicals are "anchored" on the surface at the site of a double bond. Sulfur At ordinary temperatures the stable solid form of sulfur consists of an S ring (32). In the molten form at temperatures above about 160° liquid sulfur con sists largely of long chains, which on sudden cooling to room temperature give the amorphous form of 8μ. In the vapor state the S rings break down and it seems to be fairly well established that at temperatures over 500° C. and pres sures of a few tenths of a millimeter sulfur vapor exists to the extent of over 99% as S . Rice and Sparrow (28) found that this may be frozen out on a liquid nitrogen-cooled surface to give a purple solid; on warming to room temperature it changes in a few seconds to a mixture containing about 40% of crystalline sulfur and 60% of amorphous sulfur. The activation energy of the change is 3.1 kcal. 8

8

2



6


During this work a greenish modification occasionally appeared at the edges of the purple material, and further investigation (22) showed that this green modification could be readily prepared by distilling sulfur from the molten liquid. On the Mansell system purple sulfur is characterized as 10.0 Red Purple 3/6 or 10 Red Purple 3/8 and on the National Bureau of Standards system it is charac terized as dark purplish red, 1SCC; in the same way green sulfur is characterized on the National Bureau of Standards system (15) as 7.5 Green Yellow varying from 8/4 fight yellowish green through 6/6 moderate yellowish green to 2/2 blackish green. Electron spin resonance absorption (19) has been found in both of these colored forms of sulfur and there seem to be at least two types of trapped sulfur radicals present. Red Sulfur When sulfur is subjected to gamma radiation at about —200° C. it turns a bright red but does not give any ESR absorption. When warmed to room temperature, it reverts almost explosively to ordinary yellow sulfur. At the suggestion of Charles Herzfeld we made single crystals of sulfur, both rhombic and monoclinic, obtained from saturated solutions of sulfur in toluene crystallized at appropriate temperatures. The crystals were placed in borosilicate glass tubes which were evacuated and sealed at both ends. The tubes containing the monoclinic crystals were kept at liquid nitrogen temperatures to prevent transition to the rhombic form, which is stable at room temperature. These tubes were then subjected to gamma radiation at 0.5 megaroentgen per hour for 87.2 hours. An attempted electron spin resonance determination of the radical content yielded no evidence of the presence of radicals. Sulfur Formed in Electrical Discharge Decomposition of H S 2

In these experiments H S in a flowing system at a pressure of a few tenths of a millimeter was decomposed electrically and the products were brought in contact with a liquid nitrogen-cooled finger. We used two kinds of electrical dis charge: a Raytheon microtherm unit generating radio-frequency energy at a wave length of 12.2 cm., and another obtained by connecting a voltage of 15,000 volts (maximum) to two aluminum electrodes in the gas stream. We obtained only ordinary yellow sulfur on the coldfinger,except when the power in each case was at a minimum. In the micro discharge experiments, if we kept the discharge as close as possible to the coldfingerand the power output at 50% (or lower) of the maximum, we obtained a green deposit which had a dark ring around it. On warming to room temperature, all these deposits reverted to ordinary yellow sulfur. With the electrical discharge experiment both purple and green sulfur were repeatedly obtained according to conditions. When purple sulfur was obtained, on warming a transition to green sulfur occurred at about —150° C. and finally ordinary yellow sulfur was obtained on warming to room temperature. Mixing the H S with helium or argon made no appreciable difference in our results. 2

2

Divalent Radicals There are three simple divalent radicals—CH , NH, and O—and since two of them are inorganic it seems appropriate to discuss them here. Even though most of the work has been done on methylene, it seems likely that the type of reaction characteristic for C H will be the same for N H and O. Methylene was first characterized by Rice and Glasebrook (20, 26), who prepared it by the thermal 2

2


MCE

Inorganic Free Radicals

7

decomposition of diazomethane; it is a particle of short half life of only a very few hundredths of a second. It combines with tellurium to form a red polymer, (HCHTe) , which permits its identification. It can be carried in a current of bu tane, but if the temperature is raised to about 650° C , methyl radicals are formed. This behavior led the original workers erroneously to suppose that methylene be haved like the univalent radicals and picked off hydrogen atoms. Later work showed that methylene actually "insinuates" itself in between C H bonds and its action is indiscriminate (6, 9, 12, 13). The NH radical behaves like C H (21). It also "insinuates" itself between CH bonds, forming amines. Oxygen atoms may be formed by the photochemical decomposition of either N 0 or N 0 (4) and these also react with saturated hydrocarbons to form alcohols. It seems therefore to be definitely established experimentally that the simple divalent radicals in their ground state do not behave like univalent radicals, whose main reaction with saturated molecules is to remove hydrogen atoms; instead, the divalent radicals incorporate themselves into molecules. We may understand this behavior by considering an old paper by Thomson (31) on the constitution of the atom. This paper, written only eight years after Becquerel's discovery of radio activity, postulated that the chemical atom consists of a ball of positive electricity containing electrons embedded in it. Thomson addressed himself to the problem of how the electrons would arrange themselves to form a stable system. He sug gested that the electrons would arrange themselves in rings very much resembling the periods in Mendeleyev's classification of the elements. This picture is, of course, entirely wrong and was disproved in 1912 when the famous Rutherford experiment was performed. However, the Thomson atom provides a basis for the description of a chemi cal molecule, which actually consists of a smear of negative electricity having em bedded in it the nuclei of atoms. The negative cloud is, however, not uniformly distributed but has a higher density between the nuclei. The outside of the negative cloud in organic molecules consists of electron pairs located between the nuclei of hydrogen atoms and carbon atoms. If we consider the molecule of neopentane, our picture would be a particle consisting of 12 protons each carrying a single positive charge and five carbon kernels each carrying four positive charges. The whole is embedded in a sphere of negative electricity. We may now consider a bivalent radical and what happens when it collides With a molecule such as neopentane. It seems reasonable to as sume that if the bivalent radical is in the ground state, there will be incipient bond formation by interaction with one of the electron pairs of the neopentane. This can be expected to be followed by rearrangement to form a stable molecule; in deciding what arrangement to postulate we would be guided by the principle of least motion (29). In the case of neopentane this would result in movement of one of the 12 protons and formation of an alcohol, amine, or higher hydrocarbon according to whether the divalent radical is O, NH, or C H . If instead of neo pentane, containing only primary hydrogen atoms, the hydrocarbon contains also secondary or tertiary atoms, we would expect the same sort of result. However, bond strengths would not be expected to affect the picture and the relative pro portions of isomers would be determined simply by the number of the different kinds of hydrogen atoms. Furthermore, we would expect the speed of reaction of bivalent radicals with hydrocarbon molecules to be represented by w

2


2

2

2

Ci X C2 Χ η

where

represents the concentration of the bivalent radicals, C is the concentra2



8


tion of the hydrocarbon, and η is the number of electron pairs in the sphere of negative electricity that are associated with C-H bonds. I have preferred to take this number of electron pairs because those between the carbon kernels would not be expected to interact appreciably. The foregoing discussion is limited to molecules having zero or near-zero dipole moment. Reaction of a divalent radical with a dipole will be expected not only to be strongly oriented but also to proceed with a rather low activation energy. There should be none of the indiscriminate attack by hydrocarbons that occurs with paraffin. We may understand this by noting that NH has a perma nent dipole and the oxygen atom will have an induced dipole caused by the per manent dipole of the reacting molecule. The net effect of this will be to locate interaction between the molecule and divalent radical at the negative or positive end of the molecule. We may illustrate this with reference to the blue material formed by freezing the decomposition products of hydrazoic acid (25). The primary reaction presum ably may be written HN

NH + H

3

2

so that it is tempting to ascribe the blue color to the NH radical. Foner and Hudson (8) attempted to identify NH by passing the decomposition products of N H directly into a mass spectrometer. They were unable to find any trace of NH but did find diimide, N H . Before attempting to understand the thermal decomposition of hydrazoic acid, it seems well to mention an x-ray diffraction study of the blue material made by Bolz, Mauer, and Peiser (2). It appeared to be a low temperature glass, mainly ammonium azide but also containing hydrazoic acid. According to the foregoing hypothesis, after the preliminary decomposition 3

2

2

HN

NH + H

3

(Pi)

2

various reactions can occur, depending on the conditions. Suppose first the decomposition occurs in aflowingsystem (hot tube, 800° to 950° C , at low pres sures, 0.01 to 1.0 mm.) and the gases are quickly cooled to —200° C ; we may then consider the following reactions: NH + NH NH + HN

N H 2

(Qi)

2

ΗΝ—Ν—N—NH

(Q )

H N—Ν—Ν—Ν

(Q )

ΗΝ—Ν—N—NH — 2H—Ν—N

(Ri)

3

NH + HN

3

2

HsiN—Ν—Ν—Ν — H + 2N 2

Η—Ν—Ν + H N

2

3

3

(R ) 2

2

Η—Ν—Ν—Η + N

(Si)

3

We may suppose that in the hot tube, reaction Q is negligible; N H could hardly form in this way because of the difficulty of getting rid of the energy of forma tion. Reactions Q and Q are probably highly exothermic and so we may suppose that they occur readily and are followed by Reactions R and R . In the hot tube R! would be quickly followed by S thus providing for the production of diimide. On the other hand, suppose that the first decomposition is photo chemical, the hydrazoic acid being maintained at —200° C : As before, we may then consider the reactions: 2

1

2

2

3

x

2

l5

NH + NH

N H 2

2

N H + H N -> ΗΝ—Ν—N—NH 3


(Q'i) (Q'i)

RICE

Inorganic Free Radicals NH + H N

9 H N—Ν—Ν—Ν

3

(Q'i)

2

ΗΝ—Ν—Ν—Η

2Η—Ν—Ν

(R'I)

H N—Ν—Ν—Ν

H + 2N

(R't)

2

2

2

Η—Ν—Ν + H N -+ Η—Ν—Ν—Η + N 3

3

(S'i)

While it is true that the photochemical decomposition will produce a very "hot" N H , the surroundings are at —200° C . We may then expect stabilization of the N H and Reaction Q ' may be an important source of diimide; it might be ex pected, too, that at —200° C . Reactions R\ and R ' would not be important until warming up occurred. Support for these considerations is given by Papazian (17), who studied the ultraviolet and visible absorption spectra of H N , photochemically decomposed at —200° C . He postulated the primary production of N H radicals, followed by reactions in the matrix to produce diaminohydrazine ( H N = N — N = N H ) , triazene ( H N — N = N — H ) , and the radical Η—Η—N; he ascribed the blue color and para magnetism of the solid to this radical. 2


3

2

Acknowledgment The experiments with red sulfur were conducted partly with V . de Carlo and partly with J. D . Kelley. Experiments on sulfur formed in the electrical discharge composition of H S were conducted with J. D . Kelley. 2

Literature Cited (1) Bass, A. M . , Broida, H . P., eds., "Formation and Trapping of Free Radicals," p. 400, Academic Press, New York, 1960. (2) Bolz, L . J., Mauer, F . Α., Peiser, H . S., J. Chem. Phys. 30, 349 (1959). (3) Chilton, H . T. J., Porter, G., Spectrochim. Acta 16, 390 (1960). (4) Cvetanovič, R. J., Can. J. Chem. 36, 623 (1958) and other papers. (5) Dewar, J., Jones, H., Proc. Roy Soc. London A85, 574 (1911) and earlier papers. (6) Doering, W . von E., Buttery, R. G., Laughlin, R. G., Chaudhuri, N . , J. Am. Chem. Soc. 78, 3224 (1956). (7) Dyne, P. J., Ramsay, D . Α., J. Chem. Phys. 20, 1055 (1952). (8) Foner, S. N., Hudson, R. L . , Ibid., 28, 719 (1958). (9) Frey, Η. M . , J. Am. Chem. Soc. 80, 5005 (1958). (10) Hogg, M . A. P., Spice, J. E., J. Chem. Soc. 1958, 4196. (11) Ingalls, R. B., Wall, L . , J. Chem. Phys. 35, 370 (1961). (12) Kistiakowsky, G. B., Sauer, K., J. Am. Chem. Soc. 78, 5699 (1956). (13) Knox, J. H . , Trotman-Dickenson, A. F., Chem. and Ind. (London) 1957, 268. (14) Mandelcorn, L . , Chem. Revs. 59, 827 (1959). (15) Natl. Bur. Standards, Circ. 553 (1955). (16) Norman, I., Porter, G., Nature 174, 508 (1954). (17) Papazian, Η. Α., J. Chem. Phys. 32, 456 (1960). (18) Pauling, L . , "Nature of the Chemical Bond," 3rd ed., p. 343, Cornell Univ. Press, Ithaca, Ν . Y., 1960. (19) Radford, Η. E., Rice, F . O., J. Chem. Phys. 33, 774 (1960). (20) Rice, F . O.,J.Am. Chem. Soc. 61, 213 (1939). (21) Rice, F . O., Cosgrave, D., Miller, E., unpublished work. (22) Rice, F . O., Ditter, J., J. Am. Chem. Soc. 75, 6066 (1953). (23) Rice, F . O., Ditter, J., unpublished work. (24) Rice, F . O., Freamo, M . J., J. Am. Chem. Soc. 73, 5529 (1951). (25) Ibid., 75, 548 (1953). (26) Rice, F . O., Glasebrook, A. L . , Ibid., 56, 2381 (1934). (27) Rice, F . O., Luckenbach, T., unpublished work. (28) Rice, F. O., Sparrow, C.. J. Chem. Soc. 75, 848 (1953). (29) Rice, F . O., Teller, E., J. Chem. Phys. 6, 489 (1938); 7, 199 (1939). (30) Staudinger, H., Kreis, W., Helv. Chim. Acta 8, 71 (1925). (31) Thomson, J. J., Phil. Mag. 7, 237 (1904). (32) Warren, B. E., Burwell, J. T., J. Chem. Phys. 3, 6 (1935). (33) Whittle, E., Dows, D . Α., Pimentel, G. C., Ibid., 22, 1943 (1954).

RECEIVED June 3, 1962.


Review of the Properties and Reactions of Some ... - ACS Publications

Recommend Documents